--- Log opened Sun Nov 14 00:00:32 2021 00:22 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has joined #hplusroadmap 01:54 -!- CryptoDavid [uid14990@uxbridge.irccloud.com] has joined #hplusroadmap 03:59 -!- spaceangel [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has joined #hplusroadmap 04:31 -!- Llamamoe [~Llamamoe@public-gprs631618.centertel.pl] has quit [Quit: Leaving.] 06:12 -!- red_owl [~red_owl@p5b23f9cd.dip0.t-ipconnect.de] has joined #hplusroadmap 06:31 < fltrz_> does anyone (in europe) have a physical copy of the General Index? 07:12 < fenn> 5TB compressed, 38TB uncompressed, why not just download it? 07:13 < fenn> i didn't realize that's where "indentured" came from 07:18 -!- CryptoDavid [uid14990@uxbridge.irccloud.com] has quit [Quit: Connection closed for inactivity] 07:59 < darsie> What's in the General Index? 07:59 < fenn> n-grams from all the science journals 08:02 < fenn> genetic regulatory networks are really terrible computers. they can barely do a couple logic operations in series before the whole thing falls apart due to noise. 08:04 < fenn> cryptographic immunity might have a high metabolic cost (in e.g. calories per day) 08:04 < fenn> it doesn't really solve the issues with cancer or viruses either 08:16 < fenn> fltrz i'm surprised there is a microcontroller with 50 analog inputs. i can only find analog multiplexer chips up to 32 channels, e.g. ADG731 08:28 < red_owl> cant u use cascade of multiplexers? 08:29 < red_owl> might be slow? 08:35 < fenn> not sure exactly what he's doing, maybe ultrasound imaging which would be in the 100kHz range i guess 09:22 < muurkha> EEG, so it tops out around 100Hz 09:22 < muurkha> so a cascade of multiplexers should be fine 09:22 < muurkha> I wonder if you need a preamp per input channel tho? 10:52 < fltrz_> fenn: PIC24FJ128GC010 with 50 channels 10:54 < fltrz_> I got it years ago as a free sample, so I wondered if they had similar products, but the parametric search seems split over families or is just generally horrible on mobile,... I wish mfr's would simply provide machine readable datasets of their products... 10:54 < muurkha> fltrz_: I don't think a cryptographically secure immune system is possible, because cryptography can only guarantee that you *know* something, not that you are something or will or won't do something 10:56 < muurkha> I mean, clearly it could be useful to exclude cells that don't know the password, or that don't have a unique identifier, but it's not sufficient 10:57 < fltrz_> for multicellullar organisms, upon merger of sperm cell and egg cell generate new private key. from then on all cells that stem from the fertilized egg cell has the private key 10:57 < muurkha> because a malicious cell can capture an authorized cell and reprogram it to be malicious without erasing its private key, or just copy the private key into a new malicious cell 10:57 < fltrz_> fenn: re noise in gene regulatory network, thats why it would use polymers as memory 10:58 < muurkha> oh well. I don't want to distract you from the EEG project or the optical system design project, either of which would be extremely excellent to bring to fruition 10:59 < muurkha> even if it's only fruition of the "oh, well now we know why that won't work" flavor 10:59 < fltrz_> muurkha: right, Ive continued the calculations a bit in wxMaxima 10:59 < muurkha> sweet, what are the advantages of wxMaxima over Sympy or sage? 10:59 < fltrz_> I still think it should work 10:59 < muurkha> I've never met anyone who used it 11:00 < fltrz_> I think sage would be a better choice, and I was doing it in sage, until it annoyingly returns polynomials in ascending order, which is annoying to put in LaTeX for pretty printing 11:02 < fltrz_> I tried to look up, and it seems like its non-trivial to change the default ordering of polynomials, and the few search hits were asking in the opposite direction: people saying how sage emits in descending order, but them wanting it in ascending order, with others showing how to convert it to series expansion form (ascending) near trivially. I'm afraid some dumbwit made it default behaviour, so now 11:02 < fltrz_> its non-trivial to put it in descending form 11:04 < fltrz_> then theres documentation on Sage "Term Orders" with a nauseating long list multivariate term orderings, but no examples on how to convert... 11:05 < fltrz_> so I just redid the calculation in wxMaxima and it did descending, so for now Im sticking to it. but I never marry a framework 11:07 < muurkha> aha, cool 11:08 < fltrz_> I guess unifying the different domain specific math codebases into a single codebase is just loads of thankless work... 11:08 < muurkha> yeah 11:08 < fltrz_> so instead stuff gets ducktaped together, and some default wars ensue, and we end up with crippleware 11:08 < muurkha> rarely 11:10 < fltrz_> or more paranoidly, perhaps commercial math software corps intentionally introduce major usability annoyances under the guise of contributions 11:23 < muurkha> heh 11:23 < muurkha> that's not necessary. if the core project doesn't have a strong vision of usability, it will fall apart on its own without any bad intent 11:36 < fltrz_> its certainly not necessary, but default assuming absence of malice is also dangerous 11:36 < fltrz_> threat model is "everybody plays nice, right?" 11:40 -!- CryptoDavid [uid14990@uxbridge.irccloud.com] has joined #hplusroadmap 11:40 < muurkha> I don't think so 11:40 < muurkha> the threat model of open source is "if I don't like what your patch does I won't apply it" 11:43 < muurkha> except in a few cases like ZeroMQ and Wikipedia where it's "if I don't like what your patch does I'll revert it" 11:43 < muurkha> none of these assume absence of malice 11:48 < muurkha> relevant: https://danluu.com/look-stupid/ 12:26 < kanzure> fltrz_: look up "molecular ticker tape" proposals for cellular memory devices. 12:27 < fltrz_> yeah thats been discussed a lot here for years 12:27 < fltrz_> the more I think about it, it seems bizarre that there was no selective pressure for it 12:28 < muurkha> either it doesn't work or it's hard to evolve 12:28 < fltrz_> or we're really early to the party 12:31 < muurkha> well, I guess it could be happening without us knowing it, too 12:32 < fltrz_> right 12:32 < muurkha> but sure, "hard to evolve" is relative to the time span you're looking at 12:33 < fltrz_> if all civilizations that awake, assume they are neither late nor early, then the first one would assume the same and be puzzled 12:36 < fltrz_> not trying to start a "promised universe" and "mankind as the uberspecies" sect, promise 12:39 < muurkha> there are lots of things the humans have designed and built that are hard to evolve: porcelain, steel, guns, nets, electrolysis, space rockets, wind turbines 12:40 < fltrz_> i'm not sure those things are hard to evolve, but rather shitty to maintain as they don't self repair 12:40 < muurkha> contrast: knives, flowerpots, submarines, aerodynamic flight, agriculture, pastoralism, levers, chemical reactors, all of which evolvd 12:40 < muurkha> *evolved 12:41 < muurkha> self-repairing things are necessarily comprised of non-self-repairing things, so that is not a coherent argument 12:43 < fltrz_> the evolved knives (teeth), flowerpots (?), submarines (fish), aerodynamic flight (birds), argriculture (?), ... do selfrepair, if a feather is lost a new one is grown etc... 12:43 < fltrz_> or there is redundancy (a set of teeth), and reproduction for the eventual loss 12:45 < muurkha> right, shark teeth and elephant teeth are not self-repairing, but sharks and elephants are, up to a point 12:46 < fltrz_> a wind turbine may have a very fine blade, but the wind turbine is not very self-reliant, can't repair itself in any way, can't move to windier spots, and can't reproduce. so thats not exactly a fine example of peak evolutionary fitness 12:46 < muurkha> similarly, though other fish are self-repairing, fish eggs and fish scales mostly are not 12:46 < muurkha> and there are of course always limits to self-repair in an individual organism 12:46 < fltrz_> having read write molecular tape, and being able to perhaps encrypt, send, receive and decrypt the tape seems rather useful 12:47 < muurkha> aerodynamic flight is not self-repairing; birds are self-repairing (though, as you point out, their feathers mostly are not), but airplanes are not 12:47 < muurkha> the larger human economy is, however, and it repairs or replaces damaged airplanes 12:48 < fenn> albatross soaring is similar to a wind turbine 12:48 < muurkha> a little, but albatrosses still have to eat 12:48 < fenn> nets have evolved, going by your standards 12:48 < muurkha> have they? where? 12:49 < fltrz_> right, so why would natural selection be interested in things like steel, ... we may assume it is hard to evolve, but perhaps its just not that useful in the grand scale of things, perhaps its utility is mostly a temporary thing for humans 12:49 < fenn> big gulper fish mouths, jellyfish 12:49 < muurkha> are there big gulper fish mouths that are full of holes? 12:49 < fenn> uh maybe 12:50 < fenn> the gills are there, and there are teeth and spikes and stuff in the way of the holes 12:50 < muurkha> most things that have been evolved are only useful temporarily 12:50 < fltrz_> I mean only useful for a couple of generations 12:50 < muurkha> yeah, I guess a mouth full of needlelike teeth, or baleen, is kind of netlike. so I guessI was wrong about the nets 12:50 < muurkha> stomach and hair follicle cells are useful for a lot less than a couple of generations 12:51 < muurkha> a wind turbine converts wind into useful energy, which is the major thing organisms need to survive 12:51 < fltrz_> do things like seashells count as porcelain? 12:51 < muurkha> no, seashells are calcium carbonate, a different material 12:51 < fltrz_> ok, so you do mean the exact composition, not its functionality 12:51 < muurkha> I mean, with some protein in it, giving it much greater toughess than pure calcium carbonate 12:52 < fenn> do animals mark their territory with any signals besides chemical smells (poop, pee, etc) 12:52 < muurkha> but still not in the same ballpark of strength as, you know, a toilet 12:52 < fenn> like say a "NO TRESPASSING" sign 12:52 < muurkha> sure, the humans do that 12:52 < fenn> any others? 12:53 < fltrz_> it gets sort of vague, when a bird builds the nest up high, its kind of saying: no land-crawling creatures here 12:54 < fenn> that's strictly utility, not signalling 12:54 < muurkha> https://en.wikipedia.org/wiki/Territory_%28animal%29#Visual says yes, although the visual appearance of poop is one of their examples 12:55 < muurkha> also though "Several species scratch or chew trees leaving a visual mark of their territory. This is sometimes combined with rubbing on the tree which may leave tufts of fur." 12:55 < fltrz_> muurkha: also your examples are highly domain specific, whereas arbitrary read write tape seems much more general purpouse 12:55 < fenn> yeah but that could just be smelly fur 12:55 < muurkha> also mentions wolves howling 12:56 < muurkha> fltrz_: steel is a pretty general-purpose material, vastly superior to bone in many ways 12:56 < fltrz_> ah, or butterfly patterns imitating predator eyes 12:56 < muurkha> butterfly patterns don't mark territories 12:57 < fltrz_> muurkha: but we're still comparing apples and oranges somewhat: humans dont know how to build self-repairing steel, but life has evolved self-repairing bone 12:57 < fltrz_> perhaps the figures of merit engineers use are overoptimized at the expense of a wide set of figures of merit 12:58 < fenn> the thing i was getting at was whether there were any examples of encoding abstract information in the external world in biology 12:58 < fltrz_> muurkha: butterflies don't self-immolate to mark territory? 12:58 < muurkha> maybe a better comparison would be between steel and hydroxyapatite, which is the mineral bones are made of 12:58 < fltrz_> ^ joking 12:58 < fenn> there are some ant mandibles including high quantities of zinc, which are pretty strong 12:58 < fenn> it's some sort of zinc-peptide complex, not metallic zinc 12:59 < muurkha> heh, I was gonna say 12:59 < muurkha> limpet teeth are nanostructured goethite 12:59 < muurkha> (hydroxyapatite rather than the entire nanostructured protein-hydroxyapatite composite with living cells in it, I mean) 12:59 < fltrz_> I think many metals are undesirable because of oligodynamic effect. the ones that are more common like calcium, well nearly all cell's actively pump them out 12:59 < fenn> "Limpets' teeth are composed of about 80% goethite fibres of only tens of nanometers in diameter, small enough to be flaw-insensitive, which accounts for their extreme tensile strength of 3.5–6.0 GPa and elastic modulus of 120±30 GPa" 13:00 < muurkha> yeah, I wrote that 13:00 < muurkha> the text you just pasted 13:00 < fenn> cool 13:00 * fenn highfives muurkha 13:00 < muurkha> well, the oligodynamic effect is an example of a thing that could make it hard to evolve the use of a material 13:00 < fltrz_> or just render the material useless 13:00 < muurkha> I mean it was only a slight improvement over the previous version of that line in the article 13:01 < muurkha> steel isn't useless, you just have to passivate it 13:01 < fltrz_> from biological perspective its probably also expensive to rely on rare elements 13:01 < muurkha> steel is made from the commonest of elements, though 13:02 < muurkha> iron is far more abundant than calcium, phosphorus, or carbon 13:02 < fltrz_> free metals are rare 13:02 < muurkha> especially in Earth's crust 13:02 < muurkha> hydroxyapatite and DNA are also rare, that's why we have to synthesize them 13:03 < muurkha> goethite nanofibers are apparently also pretty hard to evolve, or lots of species would use them to reinforce their bones 13:03 < fltrz_> well that ones obvious, because theres always loss mechanisms, so continuation of life requires synthesizing more and more genome carrier material 13:04 < muurkha> yeah, Dawkins might describe your whole body as "genome carrier material", hydroxyapatite and all 13:05 < fltrz_> i'd disagree, he's free to define a new concept to denote whatever he wishes to denote and demonstrate the utility of the new concept. but diluting definitions so that they lose their meaning doesn't really serve us 13:06 < muurkha> it's strange that in two subsequent lines of chat you argue that hydroxyapatite is and isn't "genome carrier material" 13:06 < muurkha> bye! 13:06 < fltrz_> I argued it was genome material? 13:06 < fltrz_> If I did, i'd take it back 13:07 < fltrz_> no one is flawless 13:24 -!- spaceangel_ [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has joined #hplusroadmap 13:27 -!- spaceangel_ [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has quit [Remote host closed the connection] 13:27 -!- spaceangel_ [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has joined #hplusroadmap 13:27 -!- spaceangel [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has quit [Ping timeout: 256 seconds] 13:32 -!- red_owl [~red_owl@p5b23f9cd.dip0.t-ipconnect.de] has quit [] 13:35 -!- spaceangel [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has joined #hplusroadmap 13:37 -!- spaceangel_ [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has quit [Ping timeout: 256 seconds] 13:51 -!- fenn [~fenn@user/fenn] has quit [Quit: Reconnecting] 13:51 -!- fenn [~fenn@user/fenn] has joined #hplusroadmap 14:42 < fltrz_> is it possible to optically stimulate sensory cells of normal head hairs? 14:42 < fltrz_> I seem to recall some recent award to someone's work on the pathways of sensory cells? 15:21 -!- CryptoDavid [uid14990@uxbridge.irccloud.com] has quit [Quit: Connection closed for inactivity] 15:34 -!- spaceangel [~spaceange@ip-62-245-71-160.net.upcbroadband.cz] has quit [Remote host closed the connection] 15:45 -!- darsie [~darsie@84-113-55-200.cable.dynamic.surfer.at] has quit [Ping timeout: 256 seconds] 16:30 -!- HumanG33k [~HumanG33k@2a01:e0a:95:5d90:215:c5ff:fe68:fb04] has joined #hplusroadmap 16:46 -!- Codaraxis [~Codaraxis@user/codaraxis] has quit [Quit: Leaving] 17:06 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=6b05c0d6 Michael Folkson: Add Jeremy and Andrew on covenants >> http://diyhpl.us/diyhpluswiki/transcripts/tabconf/2021/2021-11-05-jeremy-rubin-andrew-poelstra-covenants.md 17:06 < gnusha> https://secure.diyhpl.us/cgit/diyhpluswiki/commit/?id=c2acb845 Michael Folkson: Merge pull request #237 from michaelfolkson/jeremy-andrew-covenants >> http://diyhpl.us/diyhpluswiki/ 17:08 < kanzure> .title https://www.metzdowd.com/pipermail/cryptography/2021-November/037567.html 17:08 < saxo> [Cryptography] Linux /dev/random - a new approach 17:15 < kanzure> https://www.coindesk.com/tech/2021/11/13/taproot-bitcoins-long-anticipated-upgrade-activates-this-weekend/ 17:17 < kanzure> igem 2021 has concluded https://jamboree.igem.org/results 19:20 -!- Malvolio [~Malvolio@user/malvolio] has quit [Quit: brb] 19:23 -!- Malvolio [~Malvolio@user/malvolio] has joined #hplusroadmap 20:30 < fltrz_> so I found the [x,y] to [x',y'] map after a single ideal lens: [x,y] -> [ ((f+L)*x - L^2) / (x+f-L) , f*y / (x+f-L) ] or 1/(x+f-L) * [(f+L)*x-L^2, f*y] which heavily suggests using homogenous coordinates to get rid of the common divisor, and have simple matrix math 20:32 < fltrz_> but I'll stick to algebraic, to more easily illustrate the terms that need setting to 0 21:08 < fltrz_> muurkha: good news, both for 2 and 3 lens systems: the constraint on lens positions and focal lengths of 2 (respectively 3) lenses such that x positions get mapped linearly, is the same constraint causing y lenses to get mapped linearly, so on that front only one equation needs to be met, not a system of equations 21:08 < fltrz_> *causing y coordinates to get mapped linearly 21:09 < fltrz_> so it really is some kind of special configuration of lenses, and it seems to me plausible that this can be met with any number of lenses, as long as its not 1 21:10 < fltrz_> next up, making sure that the magnification is the same in depth and height 21:11 < fltrz_> to me this is saying something about ideal lens behaviour that I was never exposed to 21:13 < fltrz_> I did not initially expect them to be the same condition for constant x and for constant y magnification, I thought I would have ended up with a system of equations. I did suspect but was not sure that that system would have real solutions, but since its the same condition, its just a simple condition on the lens positions L_i and focal lengths f_i 21:18 < fltrz_> for 2 lenses that condition is L_2 = L_1 + f_1 + f_2 21:23 < fltrz_> thats relatively easy to position, given that placing a flat mirror (HDD platter) on one side of a lens, and a photo transparency and piece of paper on either side of the optical axis on the other side of the lens should bring the photo in focus on the paper right next to it (photo and paper in the same plane orthogonal to the optical axis) 21:24 < fltrz_> the place it comes in focus is the focal point position 21:25 < fltrz_> actually this sound like the setup for ... binoculars hmpf 21:37 < muurkha> fltrz_: yeah! 22:15 < fltrz_> ok, nevertheless, I'll continue the calculations, because they *do* imply it can be used for magnifying a small object, not just a distant object 22:16 < fltrz_> but there may still pop up other impracticalities, like absurd distances between lenses 22:19 < fltrz_> I'm continuing because theres plenty of situations where "common sense optical wisdom" doesn't actually properly deliver intuition, like Talbot Effect, where the math checks out, and the effect is confirmed, but is still unexpected from pragmatic experience (even though the effect was first observed, and later explained) 22:21 < fltrz_> its also possible that binoculars made from ideal lenses don't work with thick lenses, and one would have to make "binoculars" from mirrors... 22:21 < fltrz_> or perhaps they have to be positioned with nm precision or smth 22:22 < fltrz_> yeah I should definitely add perturbation calculations to the todo list 22:25 < fltrz_> its also logical for designers of binoculars, to optimize degrees of freedom in the lenses and doublets etc for far field at expense of microscopic behavior 22:58 -!- Netsplit *.net <-> *.split quits: muurkha 22:58 -!- Netsplit over, joins: muurkha 23:04 -!- Netsplit *.net <-> *.split quits: saxo 23:04 -!- Netsplit over, joins: saxo --- Log closed Mon Nov 15 00:00:33 2021